New Relic Studies Mobile Brower Performance

[ThePointblank]

A data-crunching startup firm, New Relic, has done analysis on the fastest mobile browser, and the results are surprising to many people:

 

http://venturebeat.com/2014/02/05/unbelievable-blackberrys-mobile-browser-is-actually-the-fastest-of-any-phonetablet-platform/

 

The fastest mobile browser? It's the Blackberry browser, able to load a website in 1.55 seconds. By comparison, the next fastest brower is Mobile Opera Mini 4.2, able to load a website in 4.78 seconds. If this isn't enough, the comparison also notes that the Blackberry brower even beats desktop browers, with the fastest desktop brower being Linux Chrome 22. A quick little display diagram showing the results is below:

 

To those in the know, this isn't really surprising. Blackberry's browers are well known for their performance, with the highest ratings for HTML5 compatiblity, ahead of everyone else, no current or upcomiong brower even close:

http://html5test.com/results/mobile.html

[SirReallySam]

Interesting! Thanks for the post!

[Khraft]

I like how they mention that Linux Chrome is the fastest desktop browser yet they don't even mention Linux anywhere else.

[nicehat]

 

 

Last year, the then new BlackBerry 10 browser smoked it's Apple competition, mobile safari and internet explorer by a wide margin. Since then, all of the players have released new operating systems, overhauled browsers and provided a better web experience overall. Things have changed, but one thing definitely has not; results of new browser speed tests published recently by New Relic show that when it comes to loading web pages, BlackBerry 10 still obliterates the competition.

 

If we look at New Relic’s data, which analyzed more than 16.8 million page loads from early October-November last year, BlackBerry 10 devices loaded pages in 1.55 sec on average. The next browser than even came close (and I use that word loosely), Opera Mini 4.2, loaded pages averaging 4.78 seconds.

 

"In other words, the BlackBerry 10 browser is more than three times faster than its next-closest competitor. Apple’s Safari browser on the iPad came in at No. 3 with an average page load time of 4.91 seconds, and no other native web browser was even included in New Relic’s top-9 rankings."

 

So BB10 is the mobile speed king for now; and with the new update 10.2 on the way, I think things are looking better already for the company. 

 

 

 

http://bgr.com/2014/02/06/blackberry-10-browser-test-iphone-android/

 

[Nineshadow]

Chrome sucks.

[lachyman]

Chrome sucks.

It's pretty good if you have the RAM to run it, so it's a no go on my 512MB phone

I currently use the Dolphin browser and I'm quite satisfied.

[kuzimins]

I'm using Chrome on my Nexus 5 because it's the default browser, and it's a bit s#it. Haven't even bothered with any other browsers, because lazy...

[Murdoch]

Such a pointless comparison, Blackberry clearly has a much smaller sample size.

[T.Vengeance]

I'm using Chrome on my Nexus 5 because it's the default browser, and it's a bit s#it. Haven't even bothered with any other browsers, because lazy...

#FirstWorldProblems

 

Can't be bothered with downloading Dolphin browser.

[kuzimins]

#FirstWorldProblems

 

Can't be bothered with downloading Dolphin browser.

I'm gonna try some others today. I used Opera Classic on my previous phone, and it was great.

[Assassin]

I love Chrome but its practically useless on my Surface Pro. Somehow they haven't fixed the dpi scaling issue. So either its too blurry or its sharp and tiny making it impossible to see. 

[nicehat]

Such a pointless comparison, Blackberry clearly has a much smaller sample size.

 

care to elaborate?

[kuddlesworth9419]

Chrome for me gets 505. But that is desktop and it's the dev channel version of Chrome. I liked to test out all the new features. 

[IlDuceSaint]

For me its the opera browser, its fast, well for me it is, so give it a try.

[Murdoch]

care to elaborate?

 

The data is from real world loading times, based on 16million+ page views. The market share of Blackberry is tiny, so the volume of page loads will be tiny, you can't really compare the speeds of browsers with such different pool sizes. You'll see that the list is automatically biased towards less used browsers. Not to mention it will favour standardised hardware, hence the lack of android, not all android hardware is created equally.

 

Also, there's no mention of which website, or on which connection, is there any java or flash running? is it enabled on the browser?

 

What's the point of releasing data like that with so little information.

[Matthias2556]

It's pretty good if you have the RAM to run it, so it's a no go on my 512MB phone

I currently use the Dolphin browser and I'm quite satisfied.

i run it on my 512mb iPod. sometimes.

actually i mostly use google now. on my iPod. google now can run on Apple's older hardware, but Siri cant, lol

[Misanthrope]

So BB10 is the mobile speed king for now; and with the new update 10.2 on the way, I think things are looking better already for the company. 

 

 

[DaRk0]

It's pretty good if you have the RAM to run it, so it's a no go on my 512MB phone

I currently use the Dolphin browser and I'm quite satisfied.

This. I have 1gb of ram and 1.2 ghz quadcore and Chrome is running just fine.

[TheIceCreamMan]

Happy with my Nexus 5 

[Eroda]

For me its the opera browser, its fast, well for me it is, so give it a try.

 

yeah even on desktop i been thinking of ditching chrome and going back to opera, chrome just seems to have dropped the ball nowdays

[mr moose]

The data is from real world loading times, based on 16million+ page views. The market share of Blackberry is tiny, so the volume of page loads will be tiny, you can't really compare the speeds of browsers with such different pool sizes. You'll see that the list is automatically biased towards less used browsers. Not to mention it will favour standardised hardware, hence the lack of android, not all android hardware is created equally.

 

Also, there's no mention of which website, or on which connection, is there any java or flash running? is it enabled on the browser?

 

What's the point of releasing data like that with so little information.

how can you say there's not enough information in the source to come to a conclusion about the validity of blackberry10 but at the same time say that it is automatically biased towards less used browsers?  Either there is enough information to start drawing conclusions or there isn't.

[ThePointblank]

This thread is a repost:

 

http://linustechtips.com/main/topic/111867-new-relic-studies-mobile-brower-performance/

[LAwLz]

I would like to know how they did this study. Supporting more HTML5 features does not mean the browser is better either.

Which phones did they use to make these tests? For all we know, the Android tests could have been done on a Nexus One (the first Nexus phone, had a single core 1GHz processor).

What sites did they test on? Did they try loading from cache one or more times? Stuff like that is what I want to know.

[ThePointblank]

I would like to know how they did this study. Supporting more HTML5 features does not mean the browser is better either.

Which phones did they use to make these tests? For all we know, the Android tests could have been done on a Nexus One (the first Nexus phone, had a single core 1GHz processor).

What sites did they test on? Did they try loading from cache one or more times? Stuff like that is what I want to know.

Everything is here:

http://blog.newrelic.com/2014/02/05/infographic-browser-wars-find-dominates-year/

 

They didn't conduct the test with phones directly, they did an aggregate of the average browser speed experienced by the end users of the nearly 3 million application instances monitored by New Relic. So, this is real world performance.

[LAwLz]

Everything is here:

http://blog.newrelic.com/2014/02/05/infographic-browser-wars-find-dominates-year/

 

They didn't conduct the test with phones directly, they did an aggregate of the average browser speed experienced by the end users of the nearly 3 million application instances monitored by New Relic. So, this is real world performance.

Then the test is pretty useless. You can't compare pieces of software (i.e browsers) if you also swap out the hardware. It's like saying IE8 is better than Firefox, but the computer you tested Firefox on had a single core CPU while the IE8 computer had a quad core.

No wonder the BlackBerry browser (I assume they mean the BB10 browser when they say "BlackBerry 5.0) is on top. All devices with it has a dual core Krait CPU, while a huge amount of Android devices have single or dual core Cortex A9 chips.

The browser marketshare is most likely also bollocks. Safari expunges the web pages in the cache far quicker than most Android phones because of the low amount of RAM, so it has to fetch web pages more often than Android devices and therefore get more page hits. Not to mention the amount of people spoofing their user agents.

 

Bad test, don't take it too seriously.